The recent development of several artificial intelligence applications—such as driverless vehicles, robo-writers and computer-aided medical diagnostics—has led to concerns about the role of human workers in the future workforce. The COVID-19 pandemic has added to these concerns, as businesses may turn to new artificial intelligence technologies to perform work activities not traditionally regarded as automatable, such as social tasks. Although the risk of automation-related job transformation is typically not distributed equally across different groups of workers, the economic consequences of the COVID-19 pandemic may be far-reaching, and could affect workers across a broad range of industries (Muro, Maxim and Whiton 2020). While previous studies have estimated the share of Canadian workers at high risk of automation-related job transformation, this study is the first to examine in great detail the automation risks faced by different groups of workers.
This study adopts a methodology similar to the one developed by Frey and Osborne (2013) and Arntz, Gregory and Zierahn (2016), and applies it to the 2016 Longitudinal and International Study of Adults, Wave 3. Frey and Osborne (2013) estimated the probability that jobs held by workers could be fully automated, based largely on input from artificial intelligence experts. Arntz, Gregory and Zierahn (2016) expanded on this by estimating adjusted automation risks in a model that took into account a broad range of job tasks and worker and firm characteristics.
This study estimates the risk faced by paid Canadian workers, after accounting for tasks, and the risk faced by specific groups of workers. It is important to note that these risk estimates are largely based on the technological feasibility of automating job tasks. There are several reasons why employers may not immediately replace humans with robots, even if it is technologically feasible to do so. These reasons include financial, legal and institutional factors; shortages in complementary skills; and product demand‑side considerations. Consequently, a high risk of automation does not necessarily imply a high risk of job loss. At the very least, it could imply a certain degree of job transformation, which is the terminology used in this study.
Results suggest that, overall, 10.6% of Canadian workers were at high risk (probability of 70% or higher) of automation-related job transformation in 2016, while 29.1% were at moderate risk (probability of between 50% and 70%). Several groups had a relatively higher share of workers who were at high risk, including those who were older (55 or above), had no postsecondary credentials or postsecondary credentials in certain fields, had low literacy or numeracy proficiency, had low employment income, or were employed part time, in small firms, in certain occupations (e.g. Office support occupations), or in the manufacturing sector. One specific finding of interest is that Business, management and public administration and Health and related fields graduates faced the highest automation-related job transformation risks among postsecondary certificate and diploma holders, but they were among the groups facing the lowest risks when looking at postsecondary degree holders.
Future research could estimate the extent to which workers classified as being at risk of automation‑related job transformation were displaced from their job soon thereafter, or participated in retraining. Another avenue for follow‑up research could examine why certain groups of workers face a higher risk of automation-related job transformation. Finally, it could also be useful to investigate how artificial intelligence has resulted in emerging occupations, shifted the composition of existing occupations, or changed the tasks performed by workers.
Chosen excerpts by Job Market Monitor. Read the whole story @ The Daily — Study: Automation and job transformation in Canada: Who is at risk?